mlsanityData to 7 Oct 2026

Cross-modal knowledge distillation for action recognition

arXiv 1910.04641 ↗First posted 10 Oct 2019cs.CVcited by 44

Not on the map. Why?

Abstract (arXiv)

In this work, we address the problem how a network for action recognition that has been trained on a modality like RGB videos can be adapted to recognize actions for another modality like sequences of 3D human poses. To this end, we extract the knowledge of the trained teacher network for the source modality and transfer it to a small ensemble of student networks for the target modality. For the cross-modal knowledge distillation, we do not require any annotated data. Instead we use pairs of sequences of both modalities as supervision, which are straightforward to acquire. In contrast to previous works for knowledge distillation that use a KL-loss, we show that the cross-entropy loss together with mutual learning of a small ensemble of student networks performs better. In fact, the proposed approach for cross-modal knowledge distillation nearly achieves the accuracy of a student network trained with full supervision.

Key facts. Cited by 44 papers in the map (data through 7 Oct 2026). First posted on arXiv 10 Oct 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 9; 2022: 6; 2023: 8; 2024: 13; 2025: 3; 2026: 5.

44 papers.

Sort:

By area:

  1. 2021Learning Student-Friendly Teacher Networks for Knowledge Distillationcited by 58
  2. 2021EvDistill: Asynchronous Events to End-task Learning via Bidirectional Reconstruction-guided Cross-modal Knowledge Distillationcited by 46
  3. 2022The Modality Focusing Hypothesis: Towards Understanding Crossmodal Knowledge Distillationcited by 44
  4. 2022A Survey on Heterogeneous Federated Learningcited by 42
  5. 2023PartDistill: 3D Shape Part Segmentation by Vision-Language Model Distillationcited by 27
  6. 2021Learning from Temporal Gradient for Semi-supervised Action Recognitioncited by 26
  7. 2024EventDance: Unsupervised Source-free Cross-modal Adaptation for Event-based Object Recognitioncited by 23
  8. 2023Categories of Response-Based, Feature-Based, and Relation-Based Knowledge Distillationcited by 19
  9. 2021DnS: Distill-and-Select for Efficient and Accurate Video Indexing and Retrievalcited by 17
  10. 2022Cross-Modal Knowledge Transfer Without Task-Relevant Source Datacited by 17
  11. 2024CrossGLG: LLM Guides One-shot Skeleton-based 3D Action Recognition in a Cross-level Mannercited by 11
  12. 2024A Survey on Multimodal Wearable Sensor-based Human Action Recognitioncited by 10
32 more
  1. 2021Transferring Knowledge Distillation for Multilingual Social Event Detectioncited by 7
  2. 2023Generative Model-based Feature Knowledge Distillation for Action Recognitioncited by 7
  3. 2024Self-Adapting Large Visual-Language Models to Edge Devices across Visual Modalitiescited by 7
  4. 2022Progressive Cross-modal Knowledge Distillation for Human Action Recognitioncited by 6
  5. 2022Text-Derived Knowledge Helps Vision: A Simple Cross-modal Distillation for Video-based Action Anticipationcited by 6
  6. 2023MoMA: Momentum Contrastive Learning with Multi-head Attention-based Knowledge Distillation for Histopathology Image Analysiscited by 6
  7. 2023VideoAdviser: Video Knowledge Distillation for Multimodal Transfer Learningcited by 6
  8. 2023Segment Any Events via Weighted Adaptation of Pivotal Tokenscited by 5
  9. 2025Distilling Cross-Modal Knowledge via Feature Disentanglementcited by 5
  10. 2021Federated Action Recognition on Heterogeneous Embedded Devicescited by 4
  11. 2022Enabling All In-Edge Deep Learning: A Literature Reviewcited by 4
  12. 2023Audio Representation Learning by Distilling Video as Privileged Informationcited by 3
  13. 2024Cross-sensor self-supervised training and alignment for remote sensingcited by 3
  14. 2024Enhancing Action Recognition from Low-Quality Skeleton Data via Part-Level Knowledge Distillationcited by 2
  15. 2024C3T: Cross-modal Transfer Through Time for Sensor-based Human Activity Recognitioncited by 2
  16. 2021End-to-End Automatic Speech Recognition with Deep Mutual Learningcited by 1
  17. 2021Distillation of Human-Object Interaction Contexts for Action Recognitioncited by 1
  18. 2024Improving Dialogue Agents by Decomposing One Global Explicit Annotation with Local Implicit Multimodal Feedbackcited by 1
  19. 2024A Generalization Theory of Cross-Modality Distillation with Contrastive Learningcited by 1
  20. 2024DL-KDD: Dual-Light Knowledge Distillation for Action Recognition in the Darkcited by 1
  21. 2024EventDance++: Language-guided Unsupervised Source-free Cross-modal Adaptation for Event-based Object Recognitioncited by 1
  22. 2024Diffusion-based Data Augmentation and Knowledge Distillation with Generated Soft Labels Solving Data Scarcity Problems of SAR Oil Spill Segmentationcited by 1
  23. 2025View-aware Cross-modal Distillation for Multi-view Action Recognitioncited by 1
  24. 2021Feature-Supervised Action Modality Transfercited by 0
  25. 2023Distilled Mid-Fusion Transformer Networks for Multi-Modal Human Activity Recognitioncited by 0
  26. 2024Leveraging Topological Guidance for Improved Knowledge Distillationcited by 0
  27. 2025Information-Theoretic Criteria for Knowledge Distillation in Multimodal Learningcited by 0
  28. 2026UMSS: Towards Unsupervised Multi-modal Semantic Segmentationcited by 0
  29. 2026Channel-wise Dynamic Knowledge Distillation via Adaptive Sample Generation for Action Recognitioncited by 0
  30. 2026Mitigating Strong-Modality Collapse in Multimodal Learning via Inverted Asymmetric Fusioncited by 0
  31. 2026Freezing of Gait Prediction Under Spatial Occlusion: An IMU-Supervised Cross-Modal Distillation Approachcited by 0
  32. 2026Efficient Quantization-Aware Distillation with Cross-Modal Alignment for Edge Vision-Language Modelscited by 0